In tropical regions, heavy precipitations may lead to catastrophic flooding due to the degradation of catchments and the expansion of settlements in flood prone zones. In the current situation, where information on rainfall and exposed assets is either scant, or requires significant time to be collected, pluvial flood risk assessments are conducted using participatory tools, without any scientific support. Another option is to use satellite precipitation products, digital terrain models and satellite images at high to moderate-resolution. However, these datasets do not reach the required accuracy at the local scale. Consequently, the potential damages and the evaluation component of risk assessment are often missing. Risk evaluation is pivotal for informed decision-making, with regards to the choice of treating or accepting the risk, implementing more effective measures, and for determining the safest areas for development. We proposed an improved method for assessing the risk of pluvial floods, which merges local and scientific knowledge and is consistent with the ISO 31010 standard. The method was successfully applied in five rural settlements in Niger and can be replicated in areas where information is scarce.